Deep Learning Fine-Tuning and Alignment: SFT, LoRA, RLHF, and DPO ByJu Yeon Eum August 24, 2025September 8, 2026